My long-term goal is to run an independent laboratory at a top-tier research institution, where we will use the larval zebrafish vestibular system as a model to understand how populations of neurons coordinate their activity to produce behavior. I postulate that the developmental processes responsible for assembling an aggregate of neurons into a functional circuit will constrain the computations the circuit can perform. Therefore, as we seek to understand how populations of neurons function, we stand to learn much from their history, and the molecular-level events that organize them. The larval zebrafish vestibular system is an ideal place to test this hypothesis. Like other vertebrates, the larval zebrafish uses a highly conserved set of neurons to robustly stabilize gaze and posture in the face of external perturbations. Unlike other vertebrates, though, their vestibular system is comprised of a small number of genetically accessible neurons, each one optically accessible throughout development. I became interested in the larval zebrafish as a model system towards the end of doctoral training. I was studying the primate oculomotor system, and it had become clear that because of the large number of neurons involved, the models of population activity I had generated to explain variability in repeated eye movements were ultimately untestable. I sought a simpler system that would allow control over complete populations of neurons, and found it as a joint post-doc in the Schier and Engert labs at Harvard. I began anew, training as a molecular geneticist of behavior, working to understand the role of a particular peptide, Hypocretin, in modulating arousal. I generated a transgenic line of fish that would let me label the neurons expressing the Hypocretin receptor, in an attempt to identify the downstream targets of the peptide, and possibly the locus of behavioral modulation. The anatomy of the neurons labeled in my transgenic line implicated the vestibular system as a potential substrate, and together with my collaborators, we set out to test this hypothesis. We measured the normal zebrafish vestibuloocular reflex (VOR), traced its anatomical substrates in my line, and showed by ablation and activation that they were necessary and sufficient for behavior. Further, we showed by genetic overexpression of Hypocretin that the VOR is sensitive to this peptide. We are currently concluding a series of electrophysiologcial experiments measuring the effects of Hypocretin on central vestibular neurons. In planning a transition to run a laboratory of my own, I sought to return to the questions of circuit function that I found intractable in non-human primates. However, it became clear that my small taste of developmental biology as a member of the Schier lab had profoundly shaped the way I approached neural circuits: I found that I often wondered not just how aggregates of neurons worked together, but about what forces shaped their fate as members of a functional circuit. It became apparent that the methodological advantages of the zebrafish were uniquely suited to answer such questions, and I sought to frame ones that would allow me to continue to learn and grow as a postdoc, while setting a course for independent work. In this proposal, I begin with a working model of the central vestibular circuit in the larval zebrafish, which proposes that two functional classes of neuron, matching those found in higher vertebrates, are sufficient to explain all of the behaviors and anatomy I had observed to date. The first set of experiments aims to test models of network-level organization and computations in the central vestibular neurons responsible for the VOR. I propose a series of experiments to first anatomically identify the classes of neurons within my model, and then to use ablations and activation of individual neurons to probe their functional roles within the population. In doing so, I will directly test for higher-level network interaction that are, to date, only theoretical propositions. Specifically, I will test whether similarly tuned neurons are redundant (i.e. individually sufficient to produce behavior, but not necessary), whether they act synergistically (i.e. each necessary, but none sufficient by itself), or whether the functional architecture of the larval zebrafish VOR reflects simple linear summation. Beyond making the first direct tests of population-level computations, each of these functional architectures suggests a certain developmental arrangement, which is the subject of my second Aim: To measure and manipulate the central vestibular system during development, to uncover the way it is wired. The central vestibular system faces a deep developmental quandary: it must link peripheral afferents with directionally tuned information to the appropriate pools of motoneurons that produce stabilizing compensatory responses. Because of the difficulty in studying development of the vestibular system in vivo, we have hypotheses of how this di-synaptic coordination might occur, but no model system in which to test them. Along with a collaborator, I have generated a second transgenic line of zebrafish that allows me to color-code each neuron in the peripheral and central vestibular system (Brainbow). I propose to use this line of fish to monitor how the vestibular system comes together. To test hypotheses of trans-synaptic coordination, I will selectively lesion populations of neurons to disrupt information flow monitor the stability of color-coded axonal projections, and measure the behavioral results of my perturbations. Together, these experiments will fill deep gaps in our understanding of how the neurons responsible for the VOR assemble and function. Further, they will fill specific technical and intellectual gaps, preparing me to successfully transition from a postdoctoral trainee to an independent investigator.